CN106021695A - Design variable stratification-based motor multi-target optimization design method - Google Patents

Abstract

The invention discloses a design variable stratification-based motor multi-target optimization design method in the field of motor optimization design. The method comprises the steps of determining an optimization model, a constraint condition and comprehensive sensitivity R(ai); when R(ai) is greater than or equal to delta, putting corresponding p design variables in a layer 1, wherein delta is the precision of a division layer; when R(ai) is less than delta, putting corresponding n-p design variables in a layer 2; obtaining relationships between k optimization targets and the p design variables in the layer 1 through a response surface method, wherein k is greater than or equal to 2 and less than or equal to n, p is greater than or equal to 1 and less than m, n is the number of optimization targets, and m is the number of design variables; determining a comprehensive optimal solution of the k optimization targets in the layer 1 according to the optimization model and the constraint condition; and determining an optimal value of the n-p design variables in the layer 2 by adopting an error function. Therefore, a plurality of the optimization targets can be simultaneously optimized and are effectively traded off under the condition that a plurality of the optimization targets conflict mutually to solve the comprehensive optimal solution.

Description

Motor multi-objective optimization design of power method based on design variable layering

Technical field

The present invention relates to Motor Optimizing Design field, particularly to motor Multiobjective Intelligent Optimization Design under constraints.

Background technology

In the optimization of motor designs, the quantity of design variable and constraints is the most, meanwhile, between each design variable Being to be mutually related, traditional optimization method is it is difficult to ensure that the accuracy of optimum results.In order to solve this problem, based on setting The method of meter variable layering has been suggested, and traditional design of electrical motor method based on design variable layering is: first from design The angle that variable is associated, is layered design variable, but is only based between two design variables and influences each other, artificially The two design variable is separated, one of them variable is one layer, and another variable and remaining variable are classified as second Layer, owing to being complementary between this bilevel optimization result, when optimizing one layer, therefore, to assure that the optimization precision of another layer Higher, this adds amount of calculation virtually.For the motor that structure is complex, the quantity of interrelated variable can be relatively Many, this will inevitably cause layering difficulty and the increase of amount of calculation.General assessment motor performance quality has multiple important Performance indications, such as, high power density, high efficiency, high reliability, wide speed regulating range, low cost, low torque ripple etc., And the method that traditional design of electrical motor method all uses tradition local parameter scanning to the design variable of different levels, to motor The assessment of energy, the most single consideration cost, do not consider the design variable impact on multiple performance indications.Due to motor Between multiple performance indications the most conflicting, such as in order to obtain higher torque density, often use higher forever Magnet consumption and less air gap thickness, but higher permanent magnet consumption and less air gap thickness will also result in location torque Increase with torque pulsation.Therefore, how in the case of meeting constraints, it is thus achieved that the comprehensive optimum of multiple performance indications Solution is still that the problem that current Motor Optimizing Design field exists.

Summary of the invention

The present invention is directed to the deficiency that conventional motors Optimization Design exists, propose a kind of motor based on design variable layering many Objective optimization method for designing, is divided into different layers by design variable, is respectively adopted the design variable of different levels different excellent Change method, solves the conflicting problem of multiple optimization aim during optimizing.

For achieving the above object, the technical side that the present invention uses based on the motor multi-objective optimization design of power method that design variable is layered Case is to comprise the following steps:

A, build the design variable a of motor to be optimized₁,a₂,…a_i,a_i+1,…a_mWith optimization aim b₁,…b_q,…b_n, determine f(a_i)_min=F (a_i,b_q), m is design variable number, m >=5, and 1 ＜ i ＜ m, n is optimization aim number, n >=1,1≤q ≤n；

B, determine Optimized model f (a_i)_min, constraints G (A) and complex sensitivitg R (a_i), as R (a_i) >=δ, by corresponding p Individual design variable is placed in layer 1, as R (a_i) ＜ δ, corresponding n-p design variable is placed in layer 2；1≤p ＜ m, δ is the precision dividing layer；

C, first pass through the relation that response phase method obtains in k optimization aim and layer 1 between p design variable, 2≤k≤n, According to Optimized model f (a_i)_minDetermine the comprehensive optimal solution of k optimization aim in layer 1 with constraints G (A), then use by mistake Difference function erf (a_i)_minDetermine n-p design variable a in layer 2_p+1,…a_nOptimal value.

Further, in step B, Optimized model f (a_i)_min=F (a_i,b_q), function F represents design variable a_iWith optimization aim b_qWeight relationship；Constraints G (A)=[g₁(a_i),g₂(a_i)…g_s(a_i)]≤0, s >=1, g_s(a_i) it is single constraints, full Foot g_s(a_i)≤0；Complex sensitivitg2≤t≤k, λ_tIt is the weight coefficient of optimization aim,E(b_q/a_i) it is a_iB time constant_qMeansigma methods, V (E (b_q/a_i)) it is E (b_q/a_i) variance, V (b_q) be b_qVariance.

Further, in step C, error function erf (a_i)_min=T (f (a_i)_min), T is the error function of Optimized model；When erf(a_i)_minDuring >=ε, update this design variable a_iValue re-optimization；As erf (a_i)_minDuring ＜ ε, then design variable a_iValue For optimal value, ε is error amount.

The present invention has the beneficial effect that after using technique scheme

1, multiple optimization aim can be optimized by the present invention simultaneously, in the case of multiple optimization aim collide with each other, effectively Multiple optimization aim are weighed to ask for comprehensive optimal solution in ground.Meanwhile, the method has general applicability and multiple target global convergence Property, it is suitable for Motor Optimizing Design application.

2, design variable is divided into according to design variable by the present invention by the influence degree of motor performance index by sensitivity analysis method Different layers, can make the most separate, simultaneously will not be because of the complicated difficulty increasing layering of electric machine structure.

3, the present invention is after optimization completes, and contrasts performance indications before and after optimizing, verifies having of optimization method of the present invention Effect property and correctness, overcome the shortcoming that traditional optimal design method is tried to gather, waste time and energy during optimizing repeatedly, nothing Need to repeatedly try to gather, optimize that precision is high, the time is short, the quick careful design for motor provides general approach.

Accompanying drawing explanation

The present invention is described in detail with detailed description of the invention below in conjunction with the accompanying drawings；

Fig. 1 is the flow chart of the motor multi-objective optimization design of power method that the present invention is layered based on design variable；

Fig. 2 is the structure chart of the flux switch permanent magnet motor described in the embodiment of the present invention；

Fig. 3 is the location torque of flux switch permanent magnet motor shown in Fig. 2 ripple before using the inventive method to optimize and after optimization Shape figure；

Fig. 4 be flux switch permanent magnet motor shown in Fig. 2 output torque use the inventive method optimize before and optimization after waveform Figure.

Detailed description of the invention

See Fig. 1, the motor multi-objective optimization design of power method that the present invention is layered, first excellent according to motor based on design variable Change design requirement, after determining ad hoc structure, first according to the architectural characteristic of concrete motor self, build motor to be optimized Design variable a₁,a₂,…a_i,a_i+1,…a_m, m is the design variable number of motor, general m >=5,1 ＜ i ＜ m.Further according to The application background of motor and self design characteristics, determine optimization aim b of motor₁,…b_q,…b_n, n is optimization aim number, N >=1,1≤q≤n.

Determine concrete optimization aim b₁,…b_q,…b_nAfterwards, it is possible to determine corresponding Optimized model f (a_i)_min:

f(a_i)_min=F (a_i,b_q),

Wherein, function F represents design variable a_iWith optimization aim b_qWeight relationship.

For different types of motor, its application background and design characteristics are often different, thus concrete performance indications are also It is different.In general, evaluate and weigh that the good and bad important indicator of motor performance has high power density, efficiently (full work is flat Face is efficient), high reliability, wide speed regulating range (5 times of radixes of ＞), there is frequent starting and acceleration and deceleration ability, low cost etc..

During motor optimizes, motor typically requires and meets some constraintss G (A), such as national standard, Yong Huyao Ask and some particular requirements etc.,

G (A)=[g₁(a_i),g₂(a_i)…g_s(a_i)]≤0, s >=1,

Wherein, g_s(a_i) it is concrete single constraints, meet g_s(a_i)≤0。

Constructing design variable a₁,a₂,…a_i,a_i+1,…a_m, and determine Optimized model f (a_i)_minAfter constraints G (A), Being mutually related impact to avoid each design variable, the present invention uses sensitivity analysis method, by design variable a₁,a₂,… a_i,a_i+1,…a_mIt is divided into different layers: layer 1 and layer 2.Sensitivity analysis method is used to differentiate clearly by finite element software Go out design variable a_iTo optimization aim b_qSensitivity H (a_i),

$H\left(a_i\right)=H\left(a_i\right)=\frac{V\left(E\right(b_q/a_i\left)\right)}{V\left(b_q\right)},$

Wherein, E (b_q/a_i) it is a_iTime constant, b_qMeansigma methods, V (E (b_q/a_i)) it is E (b_q/a_i) variance, V (b_q) it is b_qSide Difference.

On this basis, when there being k optimization aim, k >=2, k≤n, then can calculate design variable a_iMesh is optimized to k Target complex sensitivitg R (a_i):

Wherein, λ_tIt it is the weight coefficient of optimization aim.

The precision dividing layer is δ, as R (a_i) >=δ, represents corresponding p (1≤p ＜ m) individual design variable a₁,…a_pTo k Optimization aim b₁,…b_kImpact is relatively big, this p design variable a₁,…a_pThe design variable being considered as important is placed in layer 1. Otherwise, as R (a_i) ＜ δ, represent corresponding n-p design variable a_p+1,…a_nTo k optimization aim b₁,…b_kAffect less, This n-p design variable a_p+1,…a_nIt is considered as that unessential design variable is placed in layer 2.

After design variable completes layering, the design variable of layer 1 and layer 2 the two different layers is progressively optimized.Layer P design variable a in 1₁,…a_pTo k optimization aim b₁,…b_kImpact is relatively big, therefore the design variable in first optimization layer 1 a₁,…a_p.P design variable a in optimization layer 1₁,…a_pTime, n-p inessential design variable a in layer 2_p+1,…a_nProtect Hold initial value constant.First, k optimization aim b is determined by response phase method₁,…b_kWith p design variable a₁,…a_pBetween Relation.By ANSYS software, input p design variable a to be optimized₁,…a_pOptimization range, by software emulation, Obtain k optimization aim b₁,…b_kWith p design variable a₁,…a_pBetween relation.Then, according to Optimized model f (a_i)_min With constraints G (A), meet Optimized model f (a_i)_minWith the requirement of constraints G (A), it is confirmed as k optimization aim b₁,…b_kComprehensive optimal solution.

P design variable a in layer 1₁,…a_pAfter having optimized, n-p design variable a in optimization layer 2_p+1,…a_n。 N-p design variable a in optimization layer 2_p+1,…a_nTime, p design variable a in layer 1₁,…a_pThe optimal value obtained keeps Constant.Due to n-p design variable a unessential in layer 2_p+1,…a_nTo k optimization aim b₁,…b_kImpact relatively Little, the present invention uses error function erf (a_i)_minDetermine n-p design variable a in layer 2_p+1,…a_nOptimal value:

erf(a_i)_min=T (f (a_i)_min), wherein, function T is the error function of Optimized model.

Error amount is ε, as error function erf (a_i)_minDuring >=ε, update this design variable a_iValue, re-optimization, until The value of error function is less than error value epsilon.As error function erf (a_i)_minDuring ＜ ε, then it is assumed that design variable a_iValue be optimal value.

During layer 1 and layer 2 optimize, optimizing a design variable a_iAfterwards, make design variable number i=i+1, And then next design variable a is optimized_i+1。

During optimization layer 1 and layer 2, it is judged that number i having optimized design variable has reached the total of design variable the most Number m.As i ＞ m, represent all design variable a₁,a₂,…a_i,a_i+1,…a_mThe most optimised, export optimal solution set.Otherwise, Again to the design variable optimization in layer 1 and layer 2.

After layer 1 and layer 2 have optimized, need to verify the correctness of optimization method, the optimal solution set exported is carried out Checking.

Embodiment

In order to clearly demonstrate the Optimization Design of the present invention and be easy to the understanding of those skilled in the art, the present invention is with one Conventional flux switch permanent magnet motor as embodiment, elaborates the motor multiple target that the present invention is layered based on design variable excellent Change method for designing.This flux switch permanent magnet motor is as in figure 2 it is shown, mainly include outer rotor 4, and inner stator 1, " V " type is forever Magnet 3, armature winding 2 in three-phase set.Inner stator 1 has 6 stator poles, and each stator poles is embedded with two one-tenth " V " types Permanent magnet 3, the purpose that permanent magnet 3 becomes " V " type to put is to enhance poly-magnetic effect, thus further increases Air gap flux density.Permanent magnet 3 circumferentially magnetizes, and the direction of magnetization of two pieces of adjacent permanent magnets 3 is contrary.It addition, it is electric Machine has 6 armature coils, and two coils connected in series of each phase connect, and constitute threephase armature winding 2, threephase armature winding 2 Order be A phase, B phase and C phase respectively, adjacent biphase between differ 120 degree.22 teeth are had on outer rotor 4, It is referred to as 22 rotor poles.Meanwhile, outer rotor 4 is only made up of stalloy, on outer rotor 4 both without permanent magnet also without winding, Simple outer rotor 4 structure ensure that the reliability of motor, also reduces electric mach difficulty.

Use the inventive method that the flux switch permanent magnet motor in Fig. 2 is optimized design, mainly include following step:

Step 1: by the width beta of rotor pole_r, the thickness g of air gap, " V " type permanent magnet external-open angle beta_vs, permanent magnet thickness β_pm、 The height h of rotor tooth_pr, the rotor tooth polar arc width beta in yoke portion_sr, stator yoke radius R_d, stator yoke polar arc width beta_is, fixed Sub-internal diameter R_siAngle of release β in " V " type permanent magnet_vyThese 10 design variables are defined as the design variable that motor is to be optimized A=[a₁,a₂,…a₁₀], wherein, the span of each design variable is min a_i≤a_i≤max a_i。min a_i, 1 ＜ i ＜ 10, max a_iRepresent design variable a respectively_iMinima and maximum.

Step 2: as a example by flux switch permanent magnet motor is applied on city vehicle in Fig. 2 is for city vehicle, high Output torque may often be such that the requirement to meet frequent acceleration and deceleration of needs, meanwhile, in order to improve the stability of system and comfortable Property, low location torque and torque pulsation are also to need to consider.Secondly, flux switch permanent magnet motor compared to traditional forever Magneto, the double-salient-pole structure design of himself uniqueness so that motor location torque and torque pulsation are of a relatively high.Therefore, Output torque, location torque and torque pulsation are chosen as optimization aim.

Output torque T_outCan be expressed as:

$\begin{array}{c}{T}_{out}=T_{pm}+T_r+T_{cog}\\ =\frac{3}{2}p({\mathrm{\Ψ}}_{pm}{I}_q+(L_d-L_q\left)I_d{I}_q\right)+T_{cog}\end{array}---\left(1\right)$

Wherein, T_pm、T_r、T_cogIt is permanent-magnet torque, reluctance torque, location torque respectively；P is number of pole-pairs；Ψ_pmIt is permanent magnetism Magnetic linkage；I_dIt it is d shaft current；I_qIt it is q shaft current；L_dIt it is d axle inductance；L_qIt it is q axle inductance.

It should be noted that in flux switch permanent magnet motor, the value of d, q axle inductance is of substantially equal.Therefore, reluctance torque Can ignore and disregard.Then, equation (1) can be abbreviated as:

T_out=T_pm+T_cog, (2)

Torque pulsation may be considered the peak-to-peak value of output torque and the ratio of output average of torque.Therefore, torque pulsation can To be defined as:

$\begin{array}{c}{T}_{ri}=\frac{(T_{\max }-T_{\min })}{T_{ave}}\×100\%\\ =\frac{(T_{pm\_\max }+T_{cog\_\max })-(T_{pm\_\min }+T_{cog\_\min })}{T_{pm\_ave}+T_{cog\_ave}}\×100\%\\ =\frac{T_{pm\_\mathrm{var}}+T_{cog\_\mathrm{var}}}{T_{pm\_ave}+T_{cog\_ave}}\×100\%\end{array}---\left(3\right)$

Wherein, T_max、T_min、T_aveIt is the output maximum of torque, minima and meansigma methods respectively；

T_{pm_max}、T_{pm_min}、T_{pm_ave}It is the maximum of permanent-magnet torque, minima and meansigma methods respectively；

T_{cog_max}、T_{cog_min}、T_{cog_ave}It is the maximum of location torque, minima and meansigma methods respectively；

T_{pm_var}、T_{cog_var}It is the peak-to-peak value of the peak-to-peak value of permanent-magnet torque, location torque respectively.

It can be seen that output torque and torque pulsation are had very important by of a relatively high location torque from formula (2) and (3) Impact.

Once optimization aim determines, relevant Optimized model f (x_i)_minJust can draw, as shown in equation (4), this optimizes mould Type f (a_i)_minMay be considered output torque, location torque and the weighted sum of torque pulsation these three optimization aim.

Optimized model f (a_i)_min:

$f{\left(a_i\right)}_{\min }={\mathrm{\λ}}_1\frac{{T^{\′}}_{out}}{T_{out}\left(a_i\right)}+{\mathrm{\λ}}_2\frac{T_{ri}\left(a_i\right)}{{T^{\′}}_{ri}}+{\mathrm{\λ}}_3\frac{T_{cog}\left(a_i\right)}{{T^{\′}}_{cog}},---\left(4\right)$

min x_i≤x_i≤maxx_i, i=1,2 ..., m, (5)

Wherein, T'_out,T'_ri、T'_cogIt is output torque, torque pulsation and the initial value of location torque respectively,

λ₁,λ₂,λ₃It is output torque, torque pulsation and the weight coefficient of location torque respectively, between them, meets equation λ₁+λ₂+λ₃=1.

Step 3: determine main constraints g of motor_i(a_i) as follows:

Stator-rotator magnetic is close: g₁(a_i)=B_sp-B_sat≤ 0 and g₂(a_i)=B_rp-B_sat≤ 0, wherein, B_sp、B_rpAnd B_satRotor Magnetic is close and saturation magnetic induction.

Stator winding current density: g₃(a_i)=J-J_max≤0；

Efficiency: g₄(a_i)=η_min-η≤0；

Output torque: g₅(a_i)=T_min-T_out≤0；

Torque pulsation: g₆(a_i)=T_ri-(T_ri)_max≤0；

Location torque: g₇(a_i)=T_cog-(T_cog)_max≤0；

By above-mentioned constraints g_i(a_i) it is converted into total constraints G (A):

G (A)=[g₁(a_i),g₂(a_i)…g₇(a_i)]≤0 (6)

Step 4: in order to identify each design variable influence degree to each optimization aim accurately, based on sensitivity analysis side The method identification design variable influence degree to optimization aim, with sensitive degree exponent H (x_i) it is expressed as:

$H\left(a_i\right)=\frac{V\left(E\right(b_q/a_i\left)\right)}{V\left(b_q\right)},---\left(7\right)$

Wherein, E (b_q/a_i) it is a_iTime constant, b_qMeansigma methods,

V(E(b_q/a_i))、V(b_q) it is E (b respectively_q/a_i) variance, b_qVariance.

In order to synthetically calculate each design variable influence degree to optimization aim, simultaneously, it is contemplated that have during optimizing Three optimization aim.Therefore, according to formula (4) and (7), it is thus achieved that complex sensitivitg equation R (x_i), it is expressed as:

R (ai)=λ₁|Ho_ut(ai)|+λ₂|H_ri(ai)|+λ₃|H_co_g(ai) |, (8)

Wherein, | Ho_ut(a_i)|、|H_ri(a_i)|、|H_co_g(a_i) | it is output torque, torque pulsation, location torque sensitive degree exponent respectively Absolute value.It should be explained that: R (a_i) value the biggest expression respective design variable is the biggest on the impact of optimization aim, otherwise, more Little.

The width beta of rotor pole is drawn by finite element software emulation_r, the thickness g of air gap, " V " type permanent magnet external-open angle beta_vs、 Permanent magnet thickness β_pmThese 4 design variables are important design variables, are placed in layer 1, remaining 6 design variables It is placed in layer 2.

Step 5: the width beta of 4 design variable rotor poles in optimization layer 1_r, the thickness g of air gap, " V " type permanent magnet External-open angle beta_vsWith permanent magnet thickness β_pm, by ANSYS software, input the optimization range of these 4 design variables, pass through Software emulation, obtains the variation relation exporting torque, torque pulsation and location torque with these 4 design variables.Then, root According to Optimized model f (a_i)_minWith constraints G (A), determine output torque, torque pulsation and the comprehensive optimal solution of location torque.

Step 6: after 4 design variable optimizations complete in layer 1, the height of 6 design variable rotor tooths in optimization layer 2 h_pr, the rotor tooth polar arc width beta in yoke portion_sr, stator yoke radius R_d, stator yoke polar arc width beta_is, diameter of stator bore R_si" V " Angle of release β in type permanent magnet_vy, the error function erf (a of these 6 design variables is simulated by finite element software_i)_minValue, root According to error function erf (a_i)_minValue determines the optimal value of design variable in layer 2 with the magnitude relationship of error value epsilon.

Error function erf (a_i)_min:

$erf{\left(a_i\right)}_{min}=\frac{|f{\left(a_i\right)}_{min}-f^{\′}{\left(a_i\right)}_{min}|}{f^{\′}{\left(a_i\right)}_{min}},---\left(9\right)$

Wherein, f'(a_i)_minIt is after optimization layer 1, the initial value of Optimized model.

Step 7: make design variable number i=i+1, is updated to next design variable.

Step 8: if i is ＞ m, represents that all design variables are the most optimised, exports optimal solution set.Otherwise, step 5 is returned to.

Step 9: after optimization completes, needs to verify the correctness of optimization method.In an embodiment of the present invention, obtaining After the optimal solution of each design variable, analyze motor optimize before and after electromagnetic performance, see Fig. 3,4.Can from Fig. 3 Going out, after optimization, the location torque of motor substantially reduces.In Fig. 4, while motor torque ripple somewhat reduces, output turns Square significantly improves.Therefore, before and after optimization, comparing result demonstrates effectiveness and the correctness of this optimization method.

Design variable is divided into two-layer by the present invention, but, (number being typically designed variable surpasses the motor complex for structure Cross 10), design variable can be divided into three layers or even more layers.On the other hand, the present invention is by sensitivity analysis method Design variable is layered, important design variable is used response phase method, unessential design variable is used one-parameter Scanning method.But, the invention is not limited in above-mentioned specific implementation, those skilled in the art can be without departing from this Under concept thereof, the method using other, but this has no effect on the flesh and blood of the present invention, and these broadly fall into the guarantor of the present invention Protect scope.

Claims (6)

1. a motor multi-objective optimization design of power method based on design variable layering, is characterized in that comprising the following steps:

A, build the design variable a of motor to be optimized₁,a₂,…a_i,a_i+1,…a_mWith optimization aim b₁,…b_q,…b_n, determine f(a_i)_min=F (a_i,b_q), m is design variable number, m >=5, and 1 ＜ i ＜ m, n is optimization aim number, n >=1,1≤q ≤n；

B, determine Optimized model f (a_i)_min, constraints G (A) and complex sensitivitg R (a_i), as R (a_i) >=δ, by corresponding p Individual design variable is placed in layer 1, as R (a_i) ＜ δ, corresponding n-p design variable is placed in layer 2；1≤p ＜ m, δ is the precision dividing layer；

C, first pass through the relation that response phase method obtains in k optimization aim and layer 1 between p design variable, 2≤k≤n, According to Optimized model f (a_i)_minDetermine the comprehensive optimal solution of k optimization aim in layer 1 with constraints G (A), then use by mistake Difference function erf (a_i)_minDetermine the optimal value of n-p design variable in layer 2.

The most according to claim 1, motor multi-objective optimization design of power method based on design variable layering, is characterized in that: step B In, Optimized model f (a_i)_min=F (a_i,b_q), function F represents design variable a_iWith optimization aim b_qWeight relationship；Constraint bar Part G (A)=[g₁(a_i),g₂(a_i)…g_s(a_i)]≤0, s >=1, g_s(a_i) it is single constraints, meet g_s(a_i)≤0；Comprehensive sensitive Degree2≤t≤k, λ_tIt is the weight coefficient of optimization aim,It is a_iB time constant_qMeansigma methods, V (E (b_q/a_i)) it is E (b_q/a_i) variance, V (b_q) it is b_qVariance.

The most according to claim 1, motor multi-objective optimization design of power method based on design variable layering, is characterized in that: step C In, error function erf (a_i)_min=T (f (a_i)_min), T is the error function of Optimized model；As erf (a_i)_minDuring >=ε, update this Design variable a_iValue re-optimization；As erf (a_i)_minDuring ＜ ε, then design variable a_iValue be optimal value, ε is error amount.

The most according to claim 1, motor multi-objective optimization design of power method based on design variable layering, is characterized in that: step C In, in optimization layer 1 during p design variable, in layer 2, n-p design variable keeps initial value constant；At optimization layer 2 During middle n-p design variable, in layer 1, acquired optimal solution keeps constant.

The most according to claim 1, motor multi-objective optimization design of power method based on design variable layering, is characterized in that: step C In, optimizing a design variable a_iAfterwards, make design variable number i=i+1, then optimize next design variable a_i+1； Judge that number i having optimized design variable has reached m the most, as i ＞ m, export optimal solution set, otherwise re-optimization design Variable.

The most according to claim 5, motor multi-objective optimization design of power method based on design variable layering, is characterized in that: at layer 1 After having optimized with layer 2, the correctness of checking optimal solution set.